Accession Number : AD0667472

Title :   AN INTRODUCTION TO LINEAR MODELING OF STATIONARY AND NON-STATIONARY TIME SERIES.

Descriptive Note : Technical rept.,

Corporate Author : WISCONSIN UNIV MADISON DEPT OF STATISTICS

Personal Author(s) : Watts,Donald G.

Report Date : AUG 1967

Pagination or Media Count : 49

Abstract : Recent work by G. E. P. Box, of the University of Wisconsin, and G. M. Jenkins, of the University of Lancaster has led to simple models for stationary and non-stationary time series. The approach used is essentially that proposed by Zadeh and Ragazzini in 1950 and 1952, but Box and Jenkins have looked at the problem from the point of view of the statistician. Thus, while Zadeh and Ragazzini were concerned with analytic reduction of random processes to white noise, Box and Jenkins are concerned with empirical reduction of time series to uncorrelated residuals. Another important aspect of Box and Jenkins' work is that of developing useful models for seasonal time series. This paper presents an introduction to parametric modeling of time series, using the Box-Jenkins approach. (Author)

Descriptors :   (*TIME SERIES ANALYSIS, MATHEMATICAL MODELS), CORRELATION TECHNIQUES, POWER SPECTRA, RANDOM VARIABLES, PROBABILITY DENSITY FUNCTIONS, STATISTICAL FUNCTIONS, SYSTEMS ENGINEERING, TRANSFER FUNCTIONS, MATHEMATICAL PREDICTION, LEAST SQUARES METHOD, PROBLEM SOLVING

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE